Predictions on when DD2 will come to Air and DD3 to Air/Gravity, and SAE levels

[dna_at]

my predictions:

MY’27:
Air: DD1
Gravity: DD2

MY’28:
Air: DD2?
Gravity: DD2 (L2++)
Midsize: DD3 (L2++)

MY’29:
Air: DD2 (L2++)
Gravity: DD2 (L2++)
Midsize: DD3 (L3)

MY '30:
Air/Gravity/Midsize: DD3 (L4)

these are mostly based off the Investor Day video, where they mention that the Gravity will reach L2++ in 2027, that the midsize generation hardware is needed for L3/L4, and that midsize will start out at L2++ in 2027 and reach L3 in 2028. what do folks think?

 

[Adnillien]

my predictions:

MY’27:
Air: DD1
Gravity: DD2

MY’28:
Air: DD2?
Gravity: DD2 (L2++)
Midsize: DD3 (L2++)

MY’29:
Air: DD2 (L2++)
Gravity: DD2 (L2++)
Midsize: DD3 (L3)

MY '30:
Air/Gravity/Midsize: DD3 (L4)

these are mostly based off the Investor Day video, where they mention that the Gravity will reach L2++ in 2027, that the midsize generation hardware is needed for L3/L4, and that midsize will start out at L2++ in 2027 and reach L3 in 2028. what do folks think?

Since there is no industry standard definition of L2++, we will have to use the somewhat fuzzy definition that Lucid has been using. L2++ is point to point hands free including city driving. For the most part, your timeline looks reasonable but I doubt the timing for L3 and L4. L3 is not that practical for highway speeds, even with the advance sensors (long range radar and long range LiDAR) becasue the time required for a person to takeover from the automated system (respond to the alert to takeover and become situationaly aware) does not allow the vehicle to be traveling 60-80 mph. The distance traveled in that takeover time is beyond the sensor view. To make it work the car will either have to slow down or pullover in a safe and controlled manner.

I think L4 is very unlikely without very high subscription costs. The current L4 (Waymo and others) require back up drivers to remotely drive when the car gets into an unknown situation. I don't think that is possible with a bunch of individual consumer cars without a very high subscription price. The L4 equivalent of Waymo, could be be called L3 with the driver being the backup when the car stops and requests assistance. This is not the L3 scenario I discussed above where the car does a high speed hand off to the driver, it is a driver takeover after the car as safely pulled to the side of the road.

 

[dna_at]

Since there is no industry standard definition of L2++, we will have to use the somewhat fuzzy definition that Lucid has been using. L2++ is point to point hands free including city driving. For the most part, your timeline looks reasonable but I doubt the timing for L3 and L4. L3 is not that practical for highway speeds, even with the advance sensors (long range radar and long range LiDAR) becasue the time required for a person to takeover from the automated system (respond to the alert to takeover and become situationaly aware) does not allow the vehicle to be traveling 60-80 mph. The distance traveled in that takeover time is beyond the sensor view. To make it work the car will either have to slow down or pullover in a safe and controlled manner.

I think L4 is very unlikely without very high subscription costs. The current L4 (Waymo and others) require back up drivers to remotely drive when the car gets into an unknown situation. I don't think that is possible with a bunch of individual consumer cars without a very high subscription price. The L4 equivalent of Waymo, could be be called L3 with the driver being the backup when the car stops and requests assistance. This is not the L3 scenario I discussed above where the car does a high speed hand off to the driver, it is a driver takeover after the car as safely pulled to the side of the road.

I see, that is helpful insight to know about the limitations of L3 and L4, thanks for explaining
1) what will be the most likely solution for L3 highway driving - advancements in sensor view distance? those safety maneuvers you mentioned? the latter seems most likely but also not ideal
2) how do the current L4 platforms handle highway speed driving? is it a combination of the above? or do they not drive at highway speeds?
3) if Lucid (or any automaker) were to deliver L3 in 2028 as claimed, realistically would it be limited to non-highway speeds initially with L2++ for highways? at least until the above is addressed?

 

[Adnillien]

I see, that is helpful insight to know about the limitations of L3 and L4, thanks for explaining
1) what will be the most likely solution for L3 highway driving - advancements in sensor view distance? those safety maneuvers you mentioned? the latter seems most likely but also not ideal
2) how do the current L4 platforms handle highway speed driving? is it a combination of the above? or do they not drive at highway speeds?
3) if Lucid (or any automaker) were to deliver L3 in 2028 as claimed, realistically would it be limited to non-highway speeds initially with L2++ for highways? at least until the above is addressed?

Waymo is driving on freeways here in Phoenix. I think L4 is easier on highways with limited access. Since L4 is geofenced, the roads are well mapped so the likelihood of getting into an unknown situation is low unless it is low speed construction or accident scenario. Yes, Waymo can safely pull to the side on the freeway.

Mercedes delivered L3 on limited roads in 2 states with a very low speed limit. It was good fro traffic jams but not true at speed highway/freeway driving.

State regulations for L3 are not uniform making L3 even more difficult to roll out nationally. I see L3 and L4 being a very difficult task with dates much later than what is shown. I do hope that I am wrong about it.